notes

python mergeRefGene.py

#refgene_good : one gene one modal one island

#refgene_good = single + multiple_oneIsland
#single: one gene one transcript in refSeq
#multiple_oneIsland: one gene multiple transcripts in refSeq, but all at a single genomic location

#multiple_multipleIsland_good : one gene multiple transcripts in refSeq, at different genomic location, but all have the exact exon structures, can be used to map genomic data to a uniq exon annotations
#bad_gene : genes with multiple transcripts in refSeq, at different genomic location, have the more than one exon structures, very difficult to map genomic data

python refgene_to_bed.py multiple_multipleIsland > output
cut -f 1,4,6,10,11 output|grep -v "_" |sort |uniq -c > stats_multiple_structure

#grep -v chr[1-9]*_ multiple_multipleIsland > tmp

python findMultiple.py